Variability and Uncertainty in Emissions Estimation. H. C. Frey, J. Zheng, S. Li, S. Bammi, and Y. Zhao, North Carolina State University
Emissions estimation is an important step in exposure assessments that include airborne pathways, whether direct or indirect. Until recently, comparatively little attention has been given to rigorous engineering evaluation of variability and uncertainty in emissions estimates. In a series of related studies, the variability and uncertainty in emissions estimates for selected criteria and hazardous air pollutants have been characterized for selected point sources, area sources, and mobile sources. This presentation will focus on general methodological aspects of the characterization of variability and uncertainty in emissions estimates, including: (1) characterization of variability in data using parametric distributions; (2) characterization of uncertainty arising from random sampling error in statistics estimated from the distributions using bootstrap methods; (3) characterization of variability and uncertainty for censored data; (4) characterization of variability and uncertainty for mixtures of distributions; and (5) characterization of variability and uncertainty when there are measurement errors in individual data. The methods will be illustrated with selected case study examples. Knowledge regarding variability and uncertainty in emissions estimates is important in properly characterizing the range of variation of emissions for use in exposure and risk analysis. In addition, efforts to improve emission inventories can be prioritized based upon identification of the most important sources of uncertainty in an inventory.
This work is funded in part by U.S. EPA STAR Grants R826766 and R826790.
Go to . . .
2001 SRA Annual Meeting Table of Contents
2001 SRA Annual Meeting Author Index
Main Abstracts Menu Page
RiskWorld Home Page